An undergraduate student at the University of Aberdeen is developing brain-inspired computer chips – which he says could dramatically reduce the energy needed to run AI.
Henry Barnes, who is working on the AI, says he is hoping to enhance AI and bring about a “new era of low-power edge computing.”
The technology, named ‘Catalyst Neuromorphic’, won the Net Zero Challenge at the Convergence Awards this week. Rather than continuously processing information, it mimics how the brain processes information, activating only when needed.
This allows devices to perform AI tasks while consuming considerably less energy.
Adam Kosterka, Executive Director of Convergence, said: “The ambition behind Catalyst Neuromorphic is remarkable, particularly from a founder at such an early stage in his career. Henry is taking on one of the defining sustainability questions created by the rapid growth of artificial intelligence and approaching it at the fundamental level of how computing itself is powered.”
Henry is now working towards a physical prototype, as he seeks to translate the technology into commercially manufactured chips.
